It's a huge indicator. If you're 19 and performing and testing well, teams project you to continue to get better. Potential that all teams want. Your testing has room to improve, skill, technique, strength. All of it.
At 25, you're a finished product and are beating up on players 19-22 for the most part.
In my analytics grad program I did a study on age and the draft and overall I found a clear and consistent signal between age and Approximate Value for each position. It was almost universally linear that the younger the player, the better they were compared to their cohort (think of it like a bucket of similar players. I'm not comparing a 7th round DT to a top-10 pick).
This sample was over two decades of players and was pretty much for all aged 20-25 players. Younger = Better.
Especially when you smoothed it to show value of each pick as the draft went along. You smooth the value of picks to avoid saying a 6th round pick at the spot Tom Brady was picked is worth more than a 2nd round pick etc.
I still have that study somewhere and could share the raw pivot results. I don't think I have the paper itself (I didn't go for publishing), but I definitely still have the raw.